Small-Cap Value Funds: Then And Now

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Alchemist
Posts: 285
Joined: Sat Aug 30, 2014 6:35 am
Location: Florida

Re: RTM

Post by Alchemist » Sat Jul 13, 2019 3:44 pm

vineviz wrote:
Sat Jul 13, 2019 3:00 pm
So many problems here:

1) His post provided no evidence that is relevant to the ability of "factor models" to "predict the return of these funds.
2) Factor models are not supposed to "predict" future unconditional returns of live funds, so an ability or inability to do so is beside the point.
3) Factor models are designed to explain the performance of assets, which is something they do quite well.
4) His post provided no evidence that would support or refute a claim that the value or size "premium did not show up".
5) Both size and value were definitely "investable" for more than 36 years. Indeed, it was the observed real (out)performance of such investments that necessitated the formation of multifactor asset pricing models to begin with.
1) There is little to no daylight between DFSCX and DFSVX since they have both existed. One is SCV and the other SCB. The lack of real difference between them is important information in determining whether value actually provides a premium over blend in the SC space. How could this not be the case? Are you disputing that one is SCB and the other SCV?

2) This is a fair criticism of my post. The academic model only posited to explain past results not predict future ones. However, factor investing advocates do predict or at least expect premiums going forward. It is why DFA makes and markets these funds to begin with. It is why investors buy them, they are taking on the additional risk with the hope of being rewarded with a premium.

3) For the period prior to publication with a long/short definition of those factors yes, I would agree.

4) DFA's SCV fund did not provide a premium over its SCB fund. No premium for going SCV vs SCB since 1993 when the SCV fund was launched. DFA's small cap fund (they even call it a 'micro cap') did not outperform the Vanguard S&P 500 index fund since 1981 when it launched 38 years ago. Where was the premium for small over large/TSM, and where is the premium of SCV over SCB? They are not there over the entire period those funds existed. What we see is some periods where one fund is outperforming the other, but ultimately a similar total return for 38 years. Or you might say, reversion to the mean....which was the title of JoMoney's post.

5) If you can point me to the factor funds that existed prior to 1981 I would be grateful. The closest I am aware of is something like this:viewtopic.php?f=10&t=283742

User avatar
vineviz
Posts: 4420
Joined: Tue May 15, 2018 1:55 pm

Re: RTM

Post by vineviz » Sat Jul 13, 2019 4:45 pm

Alchemist wrote:
Sat Jul 13, 2019 3:44 pm

1) There is little to no daylight between DFSCX and DFSVX since they have both existed.
That's a subjective evaluation, and not one I'd agree with. Yes, the cumulative average return of the two funds is similar but the correlation of monthly returns is only 0.94. Graphically, this shows up when instead of a growth chart you look at chart of active return of one fund relative to the other. This can illustrate how the forces acting on the returns are correlated or not.

Image

Alchemist wrote:
Sat Jul 13, 2019 3:44 pm
The lack of real difference between them is important information in determining whether value actually provides a premium over blend in the SC space.
This goes probably the most important problem with trying to extrapolate the performance of risk premia from the performance of funds: unless you control for (and understand) the factor exposures of the two funds then you can't really judge the factor performance. That's what the Fama-French regressions, and not cumulative growth plots, will reveal.

From a factor exposure point of view, BOTH funds are small cap value funds: both funds have exposure to market beta, size, and value factors.

However they don't, as you'd predict, have the same degree of exposure to all three. DFSCX, as a microcap fund, unsurprisingly has more exposure to the size factor than DFSVX and less exposure to value factor.

By coincidence, the average premium for both factors (size and value) has been roughly the same from 1993 to now. The value factor has returned only very slightly more than the size factor. As the active return graph above shows, there have been very few times when the two have been as balanced as they are now: if you compare the two from 1993 to mid-2000, or from mid-2007 through today, the two funds show more "daylight. But paying attention to factor regressions and correlations allows you to decompose these elements objectively without the different pieces obfuscating each other.

Unless you control for all the variables, you can't be sure what you're seeing. Or worse, you think you're seeing one thing when actually you're seeing the opposite.

Alchemist wrote:
Sat Jul 13, 2019 3:44 pm
5) If you can point me to the factor funds that existed prior to 1981 I would be grateful. The closest I am aware of is something like this:viewtopic.php?f=10&t=283742
I've identified four funds that have returns in PortfolioVisualizer that extend back in the 1970s (1971, actually) that, at least prior to 1981, had statistically significant positive loadings on multiple factors besides the market.

In other words, all four funds were small cap value funds in the 1970s.

The list may blow your mind.

• Fidelity Equity-Income (FEQIX)
• Guggenheim Large Cap Value A (SECIX)
• Invesco Comstock A (ACSTX)
• Vanguard Windsor Inv (VWNDX)

From 1971 to 1980, an equal-weighted portfolio of these four funds produced double the CAGR of the large-cap MSCI USA index. That's despite the fact that the fund performances reflect expenses and transaction costs whereas the index does not.

Image

The early success of these actively managed funds meant that their mandates (and names, in some cases) were changed at some point so they could remain open to investors. But it is precisely the kind of success that funds like this were having that led Fama, Banz, French, etc. to try to figure out why the returns of small cap value stocks were so poorly explained by the single-factor CAPM.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
abuss368
Posts: 14071
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Small-Cap Value Funds: Then And Now

Post by abuss368 » Sat Jul 13, 2019 6:32 pm

vineviz wrote:
Sat Jul 13, 2019 2:49 pm
abuss368 wrote:
Sat Jul 13, 2019 2:11 pm
I always loved Jack Bogle's words of wisdom (and always will): Don't search for the needle in the haystack, own the haystack!
Don't forget the rest of the quote:
Don't search for the needle in the haystack, own the haystack! Except for the needles listed on the London Stock Exchange: don't own those.
:o
Mr. Bogle said that? Do you have the interview or article to share?
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

Alchemist
Posts: 285
Joined: Sat Aug 30, 2014 6:35 am
Location: Florida

Re: RTM

Post by Alchemist » Sat Jul 13, 2019 6:36 pm

vineviz wrote:
Sat Jul 13, 2019 4:45 pm
That's a subjective evaluation, and not one I'd agree with. Yes, the cumulative average return of the two funds is similar but the correlation of monthly returns is only 0.94. Graphically, this shows up when instead of a growth chart you look at chart of active return of one fund relative to the other. This can illustrate how the forces acting on the returns are correlated or not.
We can certainly agree that the correlation is 0.94. Whether that is a lot or a little is, as you correctly point out, a subjective evaluation.

vineviz wrote:
Sat Jul 13, 2019 4:45 pm
By coincidence, the average premium for both factors (size and value) has been roughly the same from 1993 to now. The value factor has returned only very slightly more than the size factor. As the active return graph above shows, there have been very few times when the two have been as balanced as they are now: if you compare the two from 1993 to mid-2000, or from mid-2007 through today, the two funds show more "daylight. But paying attention to factor regressions and correlations allows you to decompose these elements objectively without the different pieces obfuscating each other.

Unless you control for all the variables, you can't be sure what you're seeing. Or worse, you think you're seeing one thing when actually you're seeing the opposite.
My question is, is that a coincidence? Is the shorter or the longer time frames more important for determining a coincidence? From inception, 26 years, is a long time for a coincidence. Certainly far from impossible, but odd nonetheless. It seems that the size factor had more to do with both funds performance than the value factor since DFSVX's inception.

vineviz wrote:
Sat Jul 13, 2019 4:45 pm
I've identified four funds that have returns in PortfolioVisualizer that extend back in the 1970s (1971, actually) that, at least prior to 1981, had statistically significant positive loadings on multiple factors besides the market.

In other words, all four funds were small cap value funds in the 1970s.

The list may blow your mind.

• Fidelity Equity-Income (FEQIX)
• Guggenheim Large Cap Value A (SECIX)
• Invesco Comstock A (ACSTX)
• Vanguard Windsor Inv (VWNDX)

From 1971 to 1980, an equal-weighted portfolio of these four funds produced double the CAGR of the large-cap MSCI USA index. That's despite the fact that the fund performances reflect expenses and transaction costs whereas the index does not.

Image

The early success of these actively managed funds meant that their mandates (and names, in some cases) were changed at some point so they could remain open to investors. But it is precisely the kind of success that funds like this were having that led Fama, Banz, French, etc. to try to figure out why the returns of small cap value stocks were so poorly explained by the single-factor CAPM.
That is incredibly interesting. And I am very thankful for the history/background you provide here for the FF Four Factor research motivation.

I think there are two separate discussions that often get conflated, including by myself as evidenced in this thread. Those are the academic research of the Factor Model in explaining past behavior of stocks based on shared characteristics (the Factors) and the expectation of an expected premium continuing into the future. One of those is empiracle and objective, and demonstrated eloquently by your post. The second is what we bolgeheads as individual investors all with our own portfolios continue to debate.

The focus on DFSCX and DFSVX is mainly because these funds, unlike the active funds you point out above, were explicitly designed to capture the performance of small/micro and small value respectively. I think we agree that they captured both of those factors returns pretty well or as well as can be for a long-only portfolio. But they also were marketed with the intent to capturing not just the behavior of SCB and SCV, but performance premiums above and beyond the total market or S&P 500. For DFSCX, over its entire lifetime, it has not provided a premium for small cap over large cap. Or in other words, if DFSCX is faithfully capturing small/micro performance than since December 23rd of 1981 until now there has not been a premium of small/micro over the S&P 500 You can see where I am going when then comparing DFSVX with DFSCX. SCV did not, over the time period of DFSVX's inception, provide a premium to investors over SCB. An investor in DFSCX got, more or less, the same end result as an investor in DFSVX since inception. And perhaps most importantly an investor in DFSCX got, more or less, the same returns but at higher risk and cost as an investor in VFIAX.

There is of course one more implicit possibility here. Had DFSVX existed in 1981, with a 0.94 correlation with DFSCX since inception, would it have provided a premium above DFSCX of VFIAX? We cannot know of course since it did not exist for more than a decade later. But the correlation would seem to say that its returns would not deviate very far from DFSCX whether that was with increased or decreased performance.

User avatar
abuss368
Posts: 14071
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Small-Cap Value Funds: Then And Now

Post by abuss368 » Sat Jul 13, 2019 6:47 pm

Jack Bogle has an amazing speech available to read on this topic.
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
ThereAreNoGurus
Posts: 203
Joined: Fri Jan 24, 2014 11:41 pm

Re: Small-Cap Value Funds: Then And Now

Post by ThereAreNoGurus » Sat Jul 13, 2019 6:48 pm

abuss368 wrote:
Sat Jul 13, 2019 6:32 pm
vineviz wrote:
Sat Jul 13, 2019 2:49 pm
abuss368 wrote:
Sat Jul 13, 2019 2:11 pm
I always loved Jack Bogle's words of wisdom (and always will): Don't search for the needle in the haystack, own the haystack!
Don't forget the rest of the quote:
Don't search for the needle in the haystack, own the haystack! Except for the needles listed on the London Stock Exchange: don't own those.
:o
Mr. Bogle said that? Do you have the interview or article to share?
I think you missed vineviz's attempt at sarcasm/irony.

User avatar
abuss368
Posts: 14071
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Small-Cap Value Funds: Then And Now

Post by abuss368 » Sat Jul 13, 2019 6:55 pm

ThereAreNoGurus wrote:
Sat Jul 13, 2019 6:48 pm
abuss368 wrote:
Sat Jul 13, 2019 6:32 pm
vineviz wrote:
Sat Jul 13, 2019 2:49 pm
abuss368 wrote:
Sat Jul 13, 2019 2:11 pm
I always loved Jack Bogle's words of wisdom (and always will): Don't search for the needle in the haystack, own the haystack!
Don't forget the rest of the quote:
Don't search for the needle in the haystack, own the haystack! Except for the needles listed on the London Stock Exchange: don't own those.
:o
Mr. Bogle said that? Do you have the interview or article to share?
I think you missed vineviz's attempt at sarcasm/irony.
ha!
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
abuss368
Posts: 14071
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Small-Cap Value Funds: Then And Now

Post by abuss368 » Sat Jul 13, 2019 6:56 pm

ThereAreNoGurus wrote:
Sat Jul 13, 2019 6:48 pm
abuss368 wrote:
Sat Jul 13, 2019 6:32 pm
vineviz wrote:
Sat Jul 13, 2019 2:49 pm
abuss368 wrote:
Sat Jul 13, 2019 2:11 pm
I always loved Jack Bogle's words of wisdom (and always will): Don't search for the needle in the haystack, own the haystack!
Don't forget the rest of the quote:
Don't search for the needle in the haystack, own the haystack! Except for the needles listed on the London Stock Exchange: don't own those.
:o
Mr. Bogle said that? Do you have the interview or article to share?
I think you missed vineviz's attempt at sarcasm/irony.
Ha had me worried for the moment as it seemed out of Mr. Bogle's character!

Thanks!
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
ThereAreNoGurus
Posts: 203
Joined: Fri Jan 24, 2014 11:41 pm

Re: Small-Cap Value Funds: Then And Now

Post by ThereAreNoGurus » Sat Jul 13, 2019 6:59 pm

haha... you're welcome. Easy to miss sometimes with just text and a smilely. :happy

User avatar
willthrill81
Posts: 11003
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Small-Cap Value Funds: Then And Now

Post by willthrill81 » Sat Jul 13, 2019 9:34 pm

abuss368 wrote:
Sat Jul 13, 2019 6:56 pm
ThereAreNoGurus wrote:
Sat Jul 13, 2019 6:48 pm
abuss368 wrote:
Sat Jul 13, 2019 6:32 pm
vineviz wrote:
Sat Jul 13, 2019 2:49 pm
abuss368 wrote:
Sat Jul 13, 2019 2:11 pm
I always loved Jack Bogle's words of wisdom (and always will): Don't search for the needle in the haystack, own the haystack!
Don't forget the rest of the quote:
Don't search for the needle in the haystack, own the haystack! Except for the needles listed on the London Stock Exchange: don't own those.
:o
Mr. Bogle said that? Do you have the interview or article to share?
I think you missed vineviz's attempt at sarcasm/irony.
Ha had me worried for the moment as it seemed out of Mr. Bogle's character!

Thanks!
To be completely fair, there were several times when Bogle talked out of both sides of his mouth. One was 'buy the haystack but only the American one'. Another was 'don't time the market, even though I did and it worked out well for me.'
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
abuss368
Posts: 14071
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Small-Cap Value Funds: Then And Now

Post by abuss368 » Sun Jul 14, 2019 7:27 am

I would be interested to know if DFA funds have continued to increase in asset size or maybe at a smaller rate since small value is not out performing.
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
Sandtrap
Posts: 7567
Joined: Sat Nov 26, 2016 6:32 pm
Location: Hawaii No Ka Oi , N. Arizona

Re: Beating the Market?

Post by Sandtrap » Sun Jul 14, 2019 7:36 am

Taylor Larimore wrote:
Sat Jul 13, 2019 1:52 pm
Elric wrote:Nothing against indexing, owning the market, or the 3-fund portfolio (most of the stock portion of my portfolio is in an S&P index fund, and some others are legacy, as Taylor discusses), but it's equally true that indexing provides an odds on bet that an investor will underperform some other funds and that indexing is guaranteed to never be at the top of the deck.
[ quoted author corrected -- admin LadyGeek]

Bogleheads:

Never being at the bottom of the deck is more important than the risk of striving to be at the "top of the deck."
Jack Bogle's Words of Wisdom: "The simple fact is that trying to select a mutual fund that will outpace the stock market over the long term is, using Cervantes's formulation, like "looking for a needle in the haystack."
Best wishes.
Taylor
I think I need to put this on the wall in front of my desk!

The beauty of a simple "Bogle Portfolio" even with it's variations, is that it is, for the most part, "tinker proof".

Thanks, Taylor.
jim :D
Wiki Bogleheads Wiki: Everything You Need to Know

dcabler
Posts: 989
Joined: Wed Feb 19, 2014 11:30 am

Re: RTM

Post by dcabler » Sun Jul 14, 2019 7:54 am

Alchemist wrote:
Sat Jul 13, 2019 6:36 pm
vineviz wrote:
Sat Jul 13, 2019 4:45 pm
That's a subjective evaluation, and not one I'd agree with. Yes, the cumulative average return of the two funds is similar but the correlation of monthly returns is only 0.94. Graphically, this shows up when instead of a growth chart you look at chart of active return of one fund relative to the other. This can illustrate how the forces acting on the returns are correlated or not.
We can certainly agree that the correlation is 0.94. Whether that is a lot or a little is, as you correctly point out, a subjective evaluation.

vineviz wrote:
Sat Jul 13, 2019 4:45 pm
By coincidence, the average premium for both factors (size and value) has been roughly the same from 1993 to now. The value factor has returned only very slightly more than the size factor. As the active return graph above shows, there have been very few times when the two have been as balanced as they are now: if you compare the two from 1993 to mid-2000, or from mid-2007 through today, the two funds show more "daylight. But paying attention to factor regressions and correlations allows you to decompose these elements objectively without the different pieces obfuscating each other.

Unless you control for all the variables, you can't be sure what you're seeing. Or worse, you think you're seeing one thing when actually you're seeing the opposite.
My question is, is that a coincidence? Is the shorter or the longer time frames more important for determining a coincidence? From inception, 26 years, is a long time for a coincidence. Certainly far from impossible, but odd nonetheless. It seems that the size factor had more to do with both funds performance than the value factor since DFSVX's inception.

vineviz wrote:
Sat Jul 13, 2019 4:45 pm
I've identified four funds that have returns in PortfolioVisualizer that extend back in the 1970s (1971, actually) that, at least prior to 1981, had statistically significant positive loadings on multiple factors besides the market.

In other words, all four funds were small cap value funds in the 1970s.

The list may blow your mind.

• Fidelity Equity-Income (FEQIX)
• Guggenheim Large Cap Value A (SECIX)
• Invesco Comstock A (ACSTX)
• Vanguard Windsor Inv (VWNDX)

From 1971 to 1980, an equal-weighted portfolio of these four funds produced double the CAGR of the large-cap MSCI USA index. That's despite the fact that the fund performances reflect expenses and transaction costs whereas the index does not.

Image

The early success of these actively managed funds meant that their mandates (and names, in some cases) were changed at some point so they could remain open to investors. But it is precisely the kind of success that funds like this were having that led Fama, Banz, French, etc. to try to figure out why the returns of small cap value stocks were so poorly explained by the single-factor CAPM.
That is incredibly interesting. And I am very thankful for the history/background you provide here for the FF Four Factor research motivation.

I think there are two separate discussions that often get conflated, including by myself as evidenced in this thread. Those are the academic research of the Factor Model in explaining past behavior of stocks based on shared characteristics (the Factors) and the expectation of an expected premium continuing into the future. One of those is empiracle and objective, and demonstrated eloquently by your post. The second is what we bolgeheads as individual investors all with our own portfolios continue to debate.

The focus on DFSCX and DFSVX is mainly because these funds, unlike the active funds you point out above, were explicitly designed to capture the performance of small/micro and small value respectively. I think we agree that they captured both of those factors returns pretty well or as well as can be for a long-only portfolio. But they also were marketed with the intent to capturing not just the behavior of SCB and SCV, but performance premiums above and beyond the total market or S&P 500. For DFSCX, over its entire lifetime, it has not provided a premium for small cap over large cap. Or in other words, if DFSCX is faithfully capturing small/micro performance than since December 23rd of 1981 until now there has not been a premium of small/micro over the S&P 500 You can see where I am going when then comparing DFSVX with DFSCX. SCV did not, over the time period of DFSVX's inception, provide a premium to investors over SCB. An investor in DFSCX got, more or less, the same end result as an investor in DFSVX since inception. And perhaps most importantly an investor in DFSCX got, more or less, the same returns but at higher risk and cost as an investor in VFIAX.

There is of course one more implicit possibility here. Had DFSVX existed in 1981, with a 0.94 correlation with DFSCX since inception, would it have provided a premium above DFSCX of VFIAX? We cannot know of course since it did not exist for more than a decade later. But the correlation would seem to say that its returns would not deviate very far from DFSCX whether that was with increased or decreased performance.
No, but you can get an idea by looking at the data in DFA's 2019 Matrix book, which shows their index returns oftentimes back to the 1920's.
http://static.fmgsuite.com/media/docume ... ef3871.pdf

User avatar
vineviz
Posts: 4420
Joined: Tue May 15, 2018 1:55 pm

Re: RTM

Post by vineviz » Sun Jul 14, 2019 2:40 pm

Alchemist wrote:
Sat Jul 13, 2019 6:36 pm
We can certainly agree that the correlation is 0.94. Whether that is a lot or a little is, as you correctly point out, a subjective evaluation.
True. For context, though, over the same period the correlations between Vanguard Extended Market Index fund (a mid-cap index fund) and Vanguard Small Cap Index fund was 0.98.
Alchemist wrote:
Sat Jul 13, 2019 6:36 pm
My question is, is that a coincidence?
If definitely is, which is why I presented the cumulative active return graph to illustrate it. When two funds have similar factor exposures (or different exposures to highly correlated factors), that line will be much flatter. For example, here's the same cumulative active return graph for Vanguard 500 versus Vanguard Total Stock Market (notice the scaling on the Y axis).

In contrast to the two DFA funds (in which the cumulative deviation exceeded 35% on the upside and -11% on the downside), the Vanguard funds had significantly less active share (the range over the same time period was only +10% to -5%, about 2/3 less).

Image

So it's not about the shorter or longer time frames being more or less relevant, but the fact that the variance in active return is different from zero to a statistically significant degree. The graph illustrates that, but you'd need to look at the regressions to see the numbers.
Alchemist wrote:
Sat Jul 13, 2019 6:36 pm
I think there are two separate discussions that often get conflated, including by myself as evidenced in this thread. Those are the academic research of the Factor Model in explaining past behavior of stocks based on shared characteristics (the Factors) and the expectation of an expected premium continuing into the future. One of those is empiracle and objective, and demonstrated eloquently by your post. The second is what we bolgeheads as individual investors all with our own portfolios continue to debate.
I think that's totally accurate, though I'd expand the first point a little bit because there are two different points in there that also get conflated.

The conflation comes when people take a simple comparison of two funds and, without a full attribution analysis, use that comparison to make a claim about the factors. It's a conflation because the return of a fund is essentially the product of two numbers (the amount of exposure to each known factor and the performance of that factors) plus a third number, alpha (which is the residual return including things like expenses, trading costs, market impact costs, lending revenue, unknown factors, etc.).

A factor fund investor should, IMHO, insist on choosing a strategy that objectively and unequivocally execute on 2/3 of those numbers are controllable: the factor exposure and the alpha (i.e. costs). On the other hand, investors never have control over the performance of any factor: the market gives us what we get.

Which gets us to your second question: what we Bolgeheads should do with our own portfolios.

I think it's totally reasonable for investors to conclude (for whatever reason) that they are content to simplify their equity investments to a single factor like market beta and absolutely minimize possible sources of alpha (because, as Bogle eloquently told us, costs matter).

I also think it's totally reasonable for investors to diversify (for whatever reason) their equity investments across multiple robust factors.If they do so with eyes open and with an eye towards a reasonable minimization of possible sources of alpha then my personal opinion is that this is entirely consistent with Bogle's principles (even if the man himself might not approve).

These two types of investors SHOULD, imho, be able to peacefully coexist on a forum like this BUT ONLY IF each type of investor can refrain from trolling the other with misinformation and half-truths.

Investors who chase exposure to factors only because they've have recent strong performance (I'm afraid some recent momentum and/or low volatility investors might fall into the group, but probably not recent SCV investors) and/or investors who have expectations of positive portfolio alpha (i.e. active investors) have, on the other hand, drifted into oncoming traffic I fear. Discerning motive is hard on an advice forum, but I think even the most staunch factor investors would agree that we should do our best to adjust the expectations of these folks and coax them back onto the sidewalk.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
nedsaid
Posts: 11886
Joined: Fri Nov 23, 2012 12:33 pm

Re: Small-Cap Value Funds: Then And Now

Post by nedsaid » Sun Jul 14, 2019 5:56 pm

I would never put 100% of my stocks in Small Value. My belief is that you can't ignore the Large Cap stocks. Not an advocate of extreme tilts.
A fool and his money are good for business.

fennewaldaj
Posts: 742
Joined: Sun Oct 22, 2017 11:30 pm

Re: Small-Cap Value Funds: Then And Now

Post by fennewaldaj » Sun Jul 14, 2019 10:08 pm

nedsaid wrote:
Sun Jul 14, 2019 5:56 pm
I would never put 100% of my stocks in Small Value. My belief is that you can't ignore the Large Cap stocks. Not an advocate of extreme tilts.
Yeah I don't think I could ever go there either. I actually only have 3/8 of my US stocks in large caps but I really can't wrap my head around not owning them at all.

User avatar
nedsaid
Posts: 11886
Joined: Fri Nov 23, 2012 12:33 pm

Re: Small-Cap Value Funds: Then And Now

Post by nedsaid » Sun Jul 14, 2019 11:54 pm

fennewaldaj wrote:
Sun Jul 14, 2019 10:08 pm
nedsaid wrote:
Sun Jul 14, 2019 5:56 pm
I would never put 100% of my stocks in Small Value. My belief is that you can't ignore the Large Cap stocks. Not an advocate of extreme tilts.
Yeah I don't think I could ever go there either. I actually only have 3/8 of my US stocks in large caps but I really can't wrap my head around not owning them at all.
The other problem is that you could have too much tracking error when Small Cap Value hits a tough patch. Folks have a way of bailing out if they get too discouraged. Hard to say how much tilt is too much and how little is too little.
A fool and his money are good for business.

User avatar
matjen
Posts: 2087
Joined: Sun Nov 20, 2011 11:30 pm

Re: RTM

Post by matjen » Mon Jul 15, 2019 3:32 pm

vineviz wrote:
Sun Jul 14, 2019 2:40 pm

Which gets us to your second question: what we Bolgeheads should do with our own portfolios.

I think it's totally reasonable for investors to conclude (for whatever reason) that they are content to simplify their equity investments to a single factor like market beta and absolutely minimize possible sources of alpha (because, as Bogle eloquently told us, costs matter).

I also think it's totally reasonable for investors to diversify (for whatever reason) their equity investments across multiple robust factors.If they do so with eyes open and with an eye towards a reasonable minimization of possible sources of alpha then my personal opinion is that this is entirely consistent with Bogle's principles (even if the man himself might not approve).

These two types of investors SHOULD, imho, be able to peacefully coexist on a forum like this BUT ONLY IF each type of investor can refrain from trolling the other with misinformation and half-truths.

Investors who chase exposure to factors only because they've have recent strong performance (I'm afraid some recent momentum and/or low volatility investors might fall into the group, but probably not recent SCV investors) and/or investors who have expectations of positive portfolio alpha (i.e. active investors) have, on the other hand, drifted into oncoming traffic I fear. Discerning motive is hard on an advice forum, but I think even the most staunch factor investors would agree that we should do our best to adjust the expectations of these folks and coax them back onto the sidewalk.
Great post (as usual) vineviz. I find it quite telling that Taylor or Alchemist or whomever have chosen to ignore this: viewtopic.php?f=10&t=271486#p4640387

I didn't post it to prove factor investing is better but, rather, that the sky isn't falling.
A man is rich in proportion to the number of things he can afford to let alone.

Alchemist
Posts: 285
Joined: Sat Aug 30, 2014 6:35 am
Location: Florida

Re: RTM

Post by Alchemist » Mon Jul 15, 2019 8:59 pm

vineviz wrote:
Sun Jul 14, 2019 2:40 pm
In contrast to the two DFA funds (in which the cumulative deviation exceeded 35% on the upside and -11% on the downside), the Vanguard funds had significantly less active share (the range over the same time period was only +10% to -5%, about 2/3 less).

So it's not about the shorter or longer time frames being more or less relevant, but the fact that the variance in active return is different from zero to a statistically significant degree. The graph illustrates that, but you'd need to look at the regressions to see the numbers.
I of course have no disagreement on the math and my description of 'little to no daylight' was not a good one as even on the Log chart you can see ups and downs from each other. I think it is still an apt example of RTM when these two funds, like DFSCX and VFIAX, end up having more or less the same returns over a long period of time. A tell tale chart would have evened out ups/downs on performance. Yes they took different zigs and zags to get there (particularly DFSCX and VFIAX with huge differences lasting up to a decade); but for an investor the result is the same.
vineviz wrote:
Sun Jul 14, 2019 2:40 pm
A factor fund investor should, IMHO, insist on choosing a strategy that objectively and unequivocally execute on 2/3 of those numbers are controllable: the factor exposure and the alpha (i.e. costs). On the other hand, investors never have control over the performance of any factor: the market gives us what we get.

Which gets us to your second question: what we Bolgeheads should do with our own portfolios.

I think it's totally reasonable for investors to conclude (for whatever reason) that they are content to simplify their equity investments to a single factor like market beta and absolutely minimize possible sources of alpha (because, as Bogle eloquently told us, costs matter).
I think this kind of language is where Factor Investing advocates and Factor Investing skeptics start running into each other. The Factor Model is exactly that, a model, and one in a social science. But Factor Investing advocates often speak of factors as, well, facts. They are not. Not in the sense that there are facts in physics or chemistry. When someone describes TSM investing as "single factor investing" or that multi-factor investing is "more diversified" they are inherently describing another point of view from within their own framework. Yes, if you think factors drive investing returns then you are completely correct to use that terminology. But it smacks a bit condescension to non-Factor Investing advocates when you use it to describe their point of view. Unintended no doubt, but as we are all human the effect hits the same.

Indexing can be consistent with modern finance models, of course many EMH advocates index, but it is not reliant on them. This simplicity is its greatest strength. Factor investing and really any investing approach that takes its lead from academic research is subject to constant change as the research evolves. Just look at the way many advisors like Larry constantly update their recommendations or the debate over if Value is still a valid factor or explained by something else. Previous models like CAPM or MPT are subject to be overtaken by newer, better models. A classic John Bogle approach makes it much easier to stay the course as the course is not constantly being updated.

The logic of indexing? Well it stays the same. Because it is based on the basic math of how any market must work. Those of us who are generally skeptical of putting real money where social science models suggest bristle at language that assumes those models are correct. There are just too many unknown unknowns for many of us to get on board with Factor Investing (or risk parity or trend following or ect). There is nothing inherent in the market that demands any particular Factor out perform over the long haul. SCV may underperform forever, or it may provide amazing returns as it did in the early 2000's. But maybe those returns won't arrive until the 2040's. Again, nothing in the make up of the market demands it show up at any particular time or ever at all. I am not saying it will not. I am saying I do not know. The model is based on past data. That data is always being updated and influenced as people react to the previous data as they learn about it (look at all the new factor ETFs and MFs since 2000).

I do not say any of the above to try to convince you or any other investor who uses the Factor Model that they should not do so. Rather it is an attempt to explain where those of us with a different view are coming from so that there can be better communication between all of use Bogleheads and hopefully less contentious arguments as seem to be disturbingly common as of late. Also I do not hate social science, I have a graduate degree (Master of Science in International Relations) in a social science. But because social science research areas ultimately driven by the behavior of occasionally irrational and willful humans; it will never reach levels of certainty in its models the way physical sciences do. That's ok, we still learn a lot with even flawed models. It is just important to remember they all come with huge margins for error whether discussing stock returns or the way world leaders react to a security dilemma.
vineviz wrote:
Sun Jul 14, 2019 2:40 pm
I also think it's totally reasonable for investors to diversify (for whatever reason) their equity investments across multiple robust factors.If they do so with eyes open and with an eye towards a reasonable minimization of possible sources of alpha then my personal opinion is that this is entirely consistent with Bogle's principles (even if the man himself might not approve).

These two types of investors SHOULD, imho, be able to peacefully coexist on a forum like this BUT ONLY IF each type of investor can refrain from trolling the other with misinformation and half-truths.
I completely agree. I just would emphasize the eyes open portion.
vineviz wrote:
Sun Jul 14, 2019 2:40 pm
Investors who chase exposure to factors only because they've have recent strong performance (I'm afraid some recent momentum and/or low volatility investors might fall into the group, but probably not recent SCV investors) and/or investors who have expectations of positive portfolio alpha (i.e. active investors) have, on the other hand, drifted into oncoming traffic I fear. Discerning motive is hard on an advice forum, but I think even the most staunch factor investors would agree that we should do our best to adjust the expectations of these folks and coax them back onto the sidewalk.
Once again, I am in complete agreement.

I think we both agree on all of the important points. An adjustment of language, from both groups of investors, will help bring those agreements further into the light so everyone can learn from each other on the points we have different opinions on. I have already learned a great deal from this conversation with you and I appreciate your patience and effort in helping me to better understand the Factor Investing Model and its historical background.
matjen wrote:
Mon Jul 15, 2019 3:32 pm
Great post (as usual) vineviz. I find it quite telling that Taylor or Alchemist or whomever have chosen to ignore this: viewtopic.php?f=10&t=271486#p4640387

I didn't post it to prove factor investing is better but, rather, that the sky isn't falling.
I am actually a total market heretic as well and only invest in the U.S. market so international results are not something I am focused on :twisted:
Though that is all I will say for now on that as international vs no-international is a different dead horse lying half buried in other threads :wink:

JBTX
Posts: 5173
Joined: Wed Jul 26, 2017 12:46 pm

Re: Small-Cap Value Funds: Then And Now

Post by JBTX » Mon Jul 15, 2019 9:31 pm

Taylor Larimore wrote:
Sat Jul 13, 2019 2:51 am
bgreat wrote:Taylor, please, you're embarassing yourself.
bgreat:

I should have made clear that Small-Cap Value's bottom 5-year past-performance does not mean small-cap funds will continue to be at the bottom of 14 Morningstar Style Funds. I would expect (no one really knows) Small-Cap Value stocks to eventually have a period of superior performance as they return to the median (RTM).

The purpose of my post was to show the danger of factor funds seriously under-performing the market. Even if the money is not needed, no one enjoys owning a fund that needs constant re-balancing and also trails the market.

The solution is to own a simple, low-cost, diversified total market index fund which never needs rebalancing, never suffers below-market returns and is guaranteed to out-perform most investors.

What Experts Say About Total Market Index Funds
Jack Bogle's Words of Wisdom: "Indexing provides an 'odds on' bet that an investor can outpace most other equity funds, and a virtual guarantee that his performance will never be at the bottom of the deck."
Best wishes.
Taylor
I like the wording of this much better than the original post.

It is largely a behavioral issue. If you invest in total market index, and SCV wins, you'll probably shrug and chalk it up to variable returns of factors. If vice versa, you'll probably regret the decision to factor, and may end up jumping back into total market. It's one of the reasons active funds will rarely beat index funds over the long term. Any period of under performance will eventually drag the fund manager closer to a market portfolio.

Personally I've increased my allocation to SCV and international and emerging. I make no predictions but I find the underperformance a compelling reason to buy, not sell.

User avatar
willthrill81
Posts: 11003
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Small-Cap Value Funds: Then And Now

Post by willthrill81 » Mon Jul 15, 2019 9:35 pm

As Larry Swedroe recently pointed out, SCV has done a fine job over the last decade in ex-U.S. equities. DFA's DISVX has returned 6.37% vs. Vanguard's VFWIX's 4.57% from 2010 - June, 2019. We've also seen Vanguard's international small-cap fund VFSVX and iShares' international value fund EFV outperform total market ex-U.S. since 2010.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
Dialectical Investor
Posts: 510
Joined: Mon Dec 03, 2018 11:41 pm

Re: RTM

Post by Dialectical Investor » Mon Jul 15, 2019 10:13 pm

Alchemist wrote:
Mon Jul 15, 2019 8:59 pm

I think this kind of language is where Factor Investing advocates and Factor Investing skeptics start running into each other. The Factor Model is exactly that, a model, and one in a social science. But Factor Investing advocates often speak of factors as, well, facts. They are not. Not in the sense that there are facts in physics or chemistry. When someone describes TSM investing as "single factor investing" or that multi-factor investing is "more diversified" they are inherently describing another point of view from within their own framework. Yes, if you think factors drive investing returns then you are completely correct to use that terminology. But it smacks a bit condescension to non-Factor Investing advocates when you use it to describe their point of view. Unintended no doubt, but as we are all human the effect hits the same.
One reason the two camps continue to run into each other is because of the continued insistence by one group that certain facts only exist if you hold a certain viewpoint. In many cases, that is false, and it challenges the credibility of this forum. If someone wants to invest in a single factor, or advocate for it even, they should do so. There's plenty of reasons for it. Pretending they are doing something else serves no one. Speaking of only one factor doesn't even require investigating what "drives returns." The terminology is correct no matter what viewpoint you hold, and the math is what the math is whether one admits it or not. The debate should be about what one is to do consequent to that fact. The answer may be... nothing different from what one would otherwise do.

Alchemist
Posts: 285
Joined: Sat Aug 30, 2014 6:35 am
Location: Florida

Re: RTM

Post by Alchemist » Mon Jul 15, 2019 11:47 pm

Dialectical Investor wrote:
Mon Jul 15, 2019 10:13 pm
One reason the two camps continue to run into each other is because of the continued insistence by one group that certain facts only exist if you hold a certain viewpoint. In many cases, that is false, and it challenges the credibility of this forum. If someone wants to invest in a single factor, or advocate for it even, they should do so. There's plenty of reasons for it. Pretending they are doing something else serves no one. Speaking of only one factor doesn't even require investigating what "drives returns." The terminology is correct no matter what viewpoint you hold, and the math is what the math is whether one admits it or not. The debate should be about what one is to do consequent to that fact. The answer may be... nothing different from what one would otherwise do.
First of all, congrats on 500 posts! You may not have noticed, but this reply was your 500th post.

I have again used a poor choice of words. Factors are ways of categorizing groups of stocks. But why you bother characterizing them in certain ways (price to book, size, price to earnings, recent performance, etc) are arbitrary. Just look at a small cap and mid cap fund from two different providers. Where, factually, does the the cut off of small vs mid cap begin? You can arbitrarily say $10 billion or $9 billion or $13 billion. Same with value or growth. That is why the factor zoo continues to expand, more ways of grouping stocks are created and studied.

My point about 'facts' was more about the predictions some make based on the past performance of these groupings of stocks. You can say that value or large cap (however you define those) companies behaved in some way in the past. That would be a factual statement. Saying that stocks with those characteristics behave (present or future tense) in a certain way is not factual. Your prediction may be correct or incorrect, but until the future arrives it is just a prediction.

One can talk about the behavior of photons or a chemical compound in a present or future tense in a factual way because the level of certainty in models of the physical world are entirely different then levels of certainty in a model of stock behavior.

User avatar
vineviz
Posts: 4420
Joined: Tue May 15, 2018 1:55 pm

Re: RTM

Post by vineviz » Tue Jul 16, 2019 8:22 am

Alchemist wrote:
Mon Jul 15, 2019 11:47 pm
Factors are ways of categorizing groups of stocks. But why you bother characterizing them in certain ways (price to book, size, price to earnings, recent performance, etc) are arbitrary.
When discussing arbitrariness, I think it's important to distinguish between categorization of stocks (i.e. small cap vs large cap, value vs growth) and the asset pricing model used to describe stock returns (i.e. the factor models). The categories are — without question — arbitrary, whereas the factor measurements are not.

The distinction between small caps and mid caps may be arbitrary, but the S&P 600 small cap index objectively has stronger size factor loading than the CRSP mid cp index. An analogous situation in physics would be that the distinction between red light and orange light is an arbitrary one, whereas a wavelength of 655 nm is objectively longer than a wavelength of 605 nm.

When it comes to the asset pricing model, there are very robust statistical tests for whether a one asset pricing model more accurately explains the cross-sectional returns of stocks better than another asset pricing model. And we can objectively measure the impact of adding additional factors and/or replacing the previous factors in the model.

For these reasons, among others, we know that the Sharpe–Lintner CAPM (a single-factor model using just the market return) is unequivocally more explanatory than a model which posits that the cross-section of stock returns is completely random. And we know that that the Sharpe–Lintner CAPM is unequivocally less explanatory than the Fama-French three factor model (which uses market, size, and value). And so on.

In this way, using modern statistical approaches, we can proceed through the so-called "factor zoo" without becoming overly confused. Although "factor zoo" is sometimes used as a disparagement, it's not an entirely useless analogy with the field of biology. Take a look at these two butterflies.

Image

The blue butterfly is a sunburst cerulean-satyr (Caeruleuptychia helios) and the brown winged butterfly is Magneuptychia keltoumae. Two different species in two different genera. Except not: DNA testing reveals these butterflies, thought to be different species since the brown one was discovered in 1911, are just male and female forms of the same species.

DNA provides biologists a non-arbitrary way to describe living creatures, and similarly factor models provide financial economists with a non-arbitrary way to describe financial assets. This analogy shouldn't be extended too far: economics is a social, not natural, science and asset returns are FAR noisier than most data that natural scientists have to process.

Which brings me to this observation, which I think is 100% spot on.
Dialectical Investor wrote:
Mon Jul 15, 2019 10:13 pm
One reason the two camps continue to run into each other is because of the continued insistence by one group that certain facts only exist if you hold a certain viewpoint. In many cases, that is false, and it challenges the credibility of this forum. If someone wants to invest in a single factor, or advocate for it even, they should do so. There's plenty of reasons for it. Pretending they are doing something else serves no one. Speaking of only one factor doesn't even require investigating what "drives returns." The terminology is correct no matter what viewpoint you hold, and the math is what the math is whether one admits it or not. The debate should be about what one is to do consequent to that fact. The answer may be... nothing different from what one would otherwise do.
There are MANY rational reasons an investor might choose a single factor approach to stock investing using a market cap-weighted total stock market index fund. Like, some REALLY good reasons that I don't think I need to enumerate.

But there is no basis that I can see for asserting that such a fund is NOT a a single-factor fund. No matter what asset pricing model you choose, though, it is definitionally true that a total stock market index fund effectively has exposure to a single factor1.

1 Because all so-called "total market" funds use indexes which employ inclusion or investability rules, such funds frequently have economically small but statistically significant positive loadings on the "quality" factor in addition to their economically large and statistically significant positive loadings on the "market" factor.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

Alchemist
Posts: 285
Joined: Sat Aug 30, 2014 6:35 am
Location: Florida

Re: RTM

Post by Alchemist » Tue Jul 16, 2019 4:54 pm

vineviz wrote:
Tue Jul 16, 2019 8:22 am
There are MANY rational reasons an investor might choose a single factor approach to stock investing using a market cap-weighted total stock market index fund. Like, some REALLY good reasons that I don't think I need to enumerate.

But there is no basis that I can see for asserting that such a fund is NOT a a single-factor fund. No matter what asset pricing model you choose, though, it is definitionally true that a total stock market index fund effectively has exposure to a single factor1.

1 Because all so-called "total market" funds use indexes which employ inclusion or investability rules, such funds frequently have economically small but statistically significant positive loadings on the "quality" factor in addition to their economically large and statistically significant positive loadings on the "market" factor.


You may have achieved a rare feat vineviz. Changed someone's mind on the internet.

I'm now on board with describing TSM investing as single-factor and with using the language of the Factor Model for describing various subsections of the stock market since it is the most precise way we have of doing so as of now. I remain a market only investor, but will be quite happy to discuss TSM vs multi-factor investing in the context of market vs adding other factors to one's portfolio.

Perhaps we should all avoid the "Factor Investor vs non-Factor investor" language. Instead we should be focusing on market vs specific factor approaches (SCV, MOM, etc) while keeping in mind that the model describes and explains behavior of stocks on available data that necessarily comes from the past. It does not predict the future, inferring continued premiums over specific time frames (like an investor's time horizon) is separate from the academic research and model.

Post Reply